P17.03.B CYTOKINE PLASMA LEVELS AND IMMUNOPHENOTYPE IN PREOPERATIVE PERIPHERAL BLOOD IN PATIENTS WITH NEWLY DIAGNOSED GLIOBLASTOMA

Abstract BACKGROUND Glioblastoma is the deadliest type of primary brain tumor, and heterogeneity of tumor and microenvironment is one of the main challenges in defining new therapeutic strategies. We analyzed peripheral immune cell subpopulations and, after finding a protective prognostic effect of preoperative peripheral immune Natural Killer (NK) cells in a preliminary retrospective analysis, we analyzed cytokines involved in T cells and NK action in preoperative plasma samples. MATERIAL AND METHODS We analyzed preoperative peripheral blood mononuclear cells (PBMC) and related plasma cytokine levels, frozen and stored in Institutional Biobank, from 71 patients with newly diagnosed glioblastoma who underwent surgery from 2019 to 2024. Clinical and radiological features and steroid dosage were recorded at sample time. PBMC aliquotes were analyzed for immunophenotype and plasma aliquotes for simultaneous dosing of 13 cytokines through the multiplex assay LEGENDplexTM Human CD8/NK Panel (Biolegend). Acquisition was performed with BC CytoFLEX and data were processed with FCS Express 7. RESULTS Immune cells subpopulations data were available for 68 patients and cytokine levels for 67 patients. There was a great variability in cytokine levels according to what is found in general populations. All immune subpopulations and cytokines tested were not significantly influenced by steroid therapy, except for CD4+ and CD8+ T cells and plasmacytoid dendritic cell percentages. Cytokines levels were uninfluenced by age, gender and performance status. There was a direct correlation between classical monocytes and myeloid-derived suppressor cells (MDSC) and contrast-enhancing (CE) volume, while among cytokines there was a negative correlation between TNFa and CE volume. Classical monocytes were negatively correlated with TNFa, IFNg, Fas, FasL, granzyme B, and perforin. MDSC were negatively correlated with IL6, TNFa, Fas, FasL, perforin, and granzyme B. T regulatory cells were negatively correlated with TNFa, Fas, FasL, perforin, and granulysin. Mean overall survival (OS) was 15.6 months. Survival analysis (Cox regression) confirmed the results of a previous study with a positive prognostic impact of preoperative immature NK percentage and the negative impact of classical monocytes, MDSC, age, CE volume, and involvement of multiple lobes. No single cytokine had a significant impact on survival. CONCLUSION These results confirm the independent positive prognostic role of preoperative immature NK on glioblastoma patients survival. Some cytokines involved in T cells and NK action (TNFa, Fas, FasL, granzymeB, perforin) negatively correlate with classical monocytes and MDSC, which have a negative prognostic impact. TNFa is also negatively correlated with CE volume, which also has a negative independent prognostic impact.